For as long as I can remember, I’ve been a re-reader. I can’t tell you the number of times I’ve been through the Discworld canon of Terry Pratchett (whom I had the pleasure of meeting once, and a kinder, more generous soul you’ll never find). There’s a comfort, a familiarity, that revisiting these old friends provides, and with the best books, there are usually new things to discover each time around. There’s also certainty: no matter how many times I read Going Postal, the crucial semaphore message will take down the evil Grand Trunk company and the final scene will see Reacher Gilt reap his just rewards.
But today, even certainty is starting to seem a little less certain. During a discussion in my class this quarter about the ongoing evolution of the book, one of the more intriguing readings was a Quartz article by Thu-Huong Ha titled, “Are E-books Dying or Thriving? The Answer Is Yes.” The upshot of the piece, in the context of the evolution of the e-book marketplace, was that Amazon has a much better sense of what’s going on than anybody else because it controls such a substantial chunk of the market, and it collects massive amounts of data.
This all engendered a more or less predictable sense of indignation and discomfort among my students. And one passage, though perhaps peripheral to Ha’s main argument, particularly caught my students’ attention: “Only the Seattle company has deeply detailed information, down to the page, on what people want to read.”
In other words, Amazon knows not only which books and e-books people are buying but also whether or not those books are actually being read, and how they are read, down to a relatively granular level (where people stop and start, which pages they linger on, which they skip, and so on). This of course has clear business applications—for example, helping Amazon decide which authors and works to feature, or not; how to market them; and how to refine its devices and software for maximum effect. All reasonable applications, I suppose, even if tinged with Orwellian overtones. (This was the point at which I made sure my students knew about the time, back in 2009, when Amazon, in what it said was a response to a third-party rights management issue, unilaterally deleted a title from users’ Kindles—with complete lack of ironic appreciation that the title in question was Orwell’s 1984. Sometimes you have to laugh to stop yourself from crying.)
As the discussion progressed, my students began to think of what else could be done with the data Amazon collects. So many things—a targeted understanding of reading tastes and styles (oh, the research projects one could do), what’s popular, where people stop reading, and which stories, characters, or situations might cause that, as well as more sophisticated and personalized marketing approaches. But then my students dug deeper and considered the impact of machine learning, which is becoming increasingly ubiquitous and sophisticated.
Now, I have no firsthand knowledge of what Amazon is doing in this realm. But in March, Business Insider (which, deliciously, is owned by Amazon) interviewed a computer scientist named Björn Schuller, who predicted that AI could soon write better novels than humans, possibly within the next 10 years. Consider too that a real, full-blown AI might have the ability to indistinguishably replicate or mimic the style of an existing author—maybe a new Amazon-created Jane Austen novel is not so far away?
At the same time, one can imagine Amazon using the copious amounts of data it collects not just for awareness and marketing insight but to structure the reading experience itself. In other words, books may eventually be shaped by analytics and made to appeal to what readers like, while what they don’t is removed or downplayed: settings, characters, kinds of characters, scenarios, themes, language—a self-adaptive novel, if you will.
Despite the willies many readers may now be feeling, in some ways this is familiar territory. Centuries ago, as stories were shared around the fire, good storytellers would certainly have adapted and refined their stories to suit their audiences, sensing what they responded to and knowing their likes and dislikes. Theater kind of works this way, too, as does live music, or for that matter almost anything performance based. We might also be tempted to put such “adaptive” books in a category with soap operas or other serials that don’t necessarily have set endings, or choose-your-adventure books, or even video games or hypertextual works.
The creepy part comes when the charming and even heartwarming ideas of how this all works run up against our notions of fixity. We are so accustomed to novels and plays and stories that are written down, finished, done, set, that something that is at once perceptive of us and adaptive would violate our sense of stability. It would feel almost sinister somehow.
It doesn’t take long to consider the implications of books that can pander to their readers—for example, to readers’ desires for happy endings (Juliet doesn’t have to die!) or to particular points of view or cultural norms. But in this age of “fake news,” think too about what a self-adaptive nonfiction book might be like. As stories and storytelling morph, so to does what we might loosely think of as facts.
And... breathe. The reality is that the rules—and roles—of storytelling are always changing. And that’s only right. Tastes change, times change, some works endure while others fade away, perhaps to be rediscovered generations later. And besides, who knows whether any of this AI stuff is really feasible when it comes to books?
But while librarians are tooling around the 2019 ALA Midwinter in Seattle next month, in the shadow of Amazon, thinking about the future of libraries, we should consider the discussions students like mine are having in library schools these days. I just left the last session of the term for my class, in which my students were speculating on what the next generations of innovation in information resources might look like. For previous generations, even when seen by new eyes, at least the works stayed the same—the words stayed the same. But these future librarians will have to grapple with information objects of yet-unknown character, many of which will be structured so that they may constantly change and will never be done. As if the current notions of collection, organization, storage, curation, preservation, and search weren’t fraught enough.
Perhaps it’s too early to worry about how tinkering with our notions of fixity might diminish a shared set of cultural touchstones. But contemplating such a future reinforces our awareness of just how much our culture and society depend on at least some predictable, solid ground.
Joseph Janes is associate professor in the University of Washington Information School and author of Documents That Changed the Way We Live (Rowman & Littlefield)